EGU General Assembly 2020
© Author(s) 2020. This work is distributed under
the Creative Commons Attribution 4.0 License.

Subglacial Drainage Routes of the Last Scandinavian Ice Sheet

Nico Dewald1, Chris D. Clark1, Stephen J. Livingstone1, Jeremy C. Ely1, and Anna L.C. Hughes2,3
Nico Dewald et al.
  • 1Department of Geography, University of Sheffield, Sheffield, United Kingdom (
  • 2Department of Geography, University of Manchester, Manchester, United Kingdom
  • 3Department of Earth Science, University of Bergen and Bjerknes Centre for Climate Research, Bergen, Norway

The configuration of subglacial drainage systems has a major impact on the dynamics of ice sheets. However, the logistical challenges of measuring subglacial processes beneath contemporary ice sheets hinder our understanding about the spatio-temporal evolution of subglacial drainage systems. Furthermore, today’s observations on contemporary ice sheets are inherently limited to a short period within the process of deglaciation. Landforms generated by the flow of meltwater at the ice-bed interface offer the potential to study both large-scale (103-106 km2) and long-term (103-105 a) developments of subglacial drainage networks beneath past ice sheets. Despite collectively recording subglacial drainage, individual meltwater landform types such as eskers, meltwater channels and tunnel valleys, and hummock corridors have mostly been considered as separate entities. Using high-resolution (1-2 m) DEMs, we summarise the suite of interconnected subglacial meltwater landforms into a common drainage signature herein called a subglacial drainage route. Our integrated map of subglacial meltwater landforms presents the large-scale distribution of major subglacial drainage routes across Scandinavia and provides a basis for future research about the long-term evolution of subglacial drainage networks and its effect on ice dynamics of the Scandinavian Ice Sheet.

How to cite: Dewald, N., Clark, C. D., Livingstone, S. J., Ely, J. C., and Hughes, A. L. C.: Subglacial Drainage Routes of the Last Scandinavian Ice Sheet, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-17484,, 2020

Comments on the presentation

AC: Author Comment | CC: Community Comment | Report abuse

Presentation version 1 – uploaded on 05 May 2020
  • CC1: Comment on EGU2020-17484, Basile de Fleurian, 06 May 2020

    Hei Nico. Nice presentation, I am wondering if you considered the fact that different sets of landforms might have been generated at different times. From modern observation it seems unlikely that the channels would span distance greater than 50 to 100 km. Then connecting those landforms might not be the best way to proceed.

    What is your opinion on that?

    • AC1: Reply to CC1, Nico Dewald, 07 May 2020

      Hej Basile,

      you're certainly right, some of these landforms have likely been formed at very different times.

      This means, of course, our map only shows the ways (or routes) where subglacial meltwater flowed at some point during the last glaciation. This would be similar if we were only looking at the ice sheet scale distribution of eskers, for example. 

      We also plan to study the temporal relations of these landforms later on by looking at their relative positions both to each other and to ice-marginal features.

      I hope this answers your question?

  • CC2: Comment on EGU2020-17484, Mark Johnson, 08 May 2020

    Hej Nico, Nice poster and thanks for mentioning murtoos and the other work. A couple comments and a question. First, much of the bedrock topography in Sweden is old (Mesozoic weathering. see papers by Lidmar-Bergström and Olvmo). This nmeans that a number of valleys that likely are occupied by drainages are not actually made by the ice but inherited. The second comments, which I think I asked previously, is that there is a tme-transgressive nature to the fluvial features, not only acorss the ice hseet (as the ice retreats) but also as each site, as the bed conditions change. I am sure you are aware of this. A question would be, what is it that you can see in your results that, for example, one can't see in the Stoeven esker map? I assume is that by combining all, you get a gretaer density. But what new things do you see?

    • AC2: Reply to CC2, Nico Dewald, 12 May 2020

      Hej Mark,

      thanks a lot for your comments and I fully agree with them, of course. 

      I'd like to add that I think we should consider inherited valleys that played an active part in the subglacial drainage system as well if we want to get a full picture about subglacial drainage networks and understand how they work. I see the challenge in distinguishing between valleys that were an active part of the drainage system and those that were not. Widely observed associations with eskers give me confidence that also the inherited routes that I've mapped were an active part of the subglacial drainage system during the last deglaciation; there are some eskers within the 'joint-valleys' in Blekinge, too, for example. I should note that I'm assigning a confidence level to all routes based on its landforms, so I will be able to distinguish between routes that are associated with eskers (= higher confidence) and those that were not (= lower confidence) (Should have added this to the presentation, perhaps).

      Regarding your question, we can see a greater density but we can also see how it changes across the ice sheet. Further, (potential) tributary connections and their different angles become more visible (I'm planning to have a closer look at that later). Also, areas with a general lack of subglacial meltwater landforms become much more apparent. I think it would be interesting to see how all these changes relate to palaeo-ice dynamics and as I said, I think we have to include all routes of subglacial meltwater to get a better idea about how these systems work. Beyond that, I think comparisons between individual ice sheets (e.g. to Emma's mapping, which uses a very similar approach) would be interesting to see as well.

      Eventually, this is but the first part of my PhD and I want to look at some areas in a bit more detail, too. This map (and the mapping process itself) is also going to helping me choosing appropriate areas.

      Please let me know if you have any more questions!

      • CC5: Reply to AC2, Mark Johnson, 12 May 2020

        Thanks for the answers! Good luck with the project.

  • CC3: Comment on EGU2020-17484, Matteo Spagnolo, 08 May 2020

    Hi Nico,

    Very interesting.

    You might have explained this already in one of the chats that I missed, but I wonder how exactly your semplification/linking process works. The example you show is fairly simple, but I can imagine many other, more complex situations. For example, are there threshold distances/buffers applied to determine which landforms should/could be joined? Is the process done automatically? 


    • AC3: Reply to CC3, Nico Dewald, 12 May 2020

      Hi Matteo,

      I agree, the example in the presentation is fairly straightforward and there are some more difficult cases. All mapping is done manually, so there is a subjective component involved in the mapping process. Other people on PalGlac are currently working on automating some steps (e.g. hummock corridors as in Lewington et al., 2019, 'An automated method for mapping geomorphological expressions of former subglacial meltwater pathways (hummock corridors) from high resolution digital elevation data'), but the main task I'm dealing with is mapping these landforms manually.

      When it comes to joining up individual routes, I look for morphological expressions such as the "topographic step" on slide 5 and decisions whether I should join two routes or not are based on these (in cases like the one on slide 5, I'm not joining them). Consequently, if there's no morphological expression linking two features, I refrain from joining them during the mapping process. However, similar to Storrar et al. (2014, 'Morphometry and pattern of a large sample (>20,000) of Canadian eskers and implications for subglacial drainage beneath ice sheets'), apparent routes can be joined for further analysis at a later stage, too, of course.

      I hope this answers your question? Please let me know if you have any more questions or comments!

      • CC4: Reply to AC3, Matteo Spagnolo, 12 May 2020

        Thank you Nico, all clear now. Initially, I thought you were using an automatic approach, but I am glad to hear you are not.